EV Charging Compliance Retrofit: Audit Before You Buy

Compliance conversations usually start with the wrong question. Operators ask whether a charger is certified, when the more useful question is whether the unit already standing on site is still covered by the certificate they were handed. Those are two different things. A product can hold a valid approval and still be non-compliant in the field, because the meter was swapped, a card reader was bolted on, or the variant installed is not the variant that was assessed.

The audit below is written for operators with hardware already in the ground. It runs in six steps and ends with a triage decision instead of a shopping list.

Why the audit comes before the quote

A retrofit quote for an unknown fleet is a guess dressed up as a price. Before anyone can say whether a unit needs a new module or a new cabinet, they need to know how many hardware builds are on site, which controllers they run and which meters are inside them. Without that, money goes two ways at once: full replacements where a certified part would have done, and parts bolted onto a platform already past its support horizon.

The second reason is paperwork. A declaration of conformity is issued for a defined configuration. Anything added afterwards sits outside it until the declaration is reissued. That single fact reshapes most retrofit budgets, because the cheapest hardware change is often the one that creates the most documentation work.

Step one: build the register before you read a single quote

One row per unit, not one row per model. The register is the document every later decision depends on.

Fill it in from the nameplate, the label inside the door and the back-end, not from the purchase order. The purchase order records what was intended; the labels record what arrived.

Step two: test the payment path the way a stranger would

Stand at the unit without the operator's app, without a membership, without calling anyone. If you cannot start a charge and pay for it, the unit does not offer ad-hoc payment, whatever the feature list says. This is the most common gap in an installed fleet, and the one that matters most, because a payment method requiring registration excludes every driver who has not registered in advance.

Record four things per unit: whether a card reader is physically present, whether it is integrated into the cabinet or mounted beside it, whether it reads contactless cards, and who processes the transaction. That last point matters more than operators expect. A terminal owned by a third party may satisfy the driver but leave the site owner without a clear answer on refunds, receipts and settlement.

Step three: check the metering chain, not the meter reading

Billing-grade metering is a chain rather than a component: an approved meter, correct installation, an intact seal, and a record tying the meter serial to that specific unit. Break any link and the kWh figure can be perfectly accurate and still unusable for billing. Accuracy is a measurement property; accountability is a documentation property.

Is the seal intact, and is the seal number written down anywhere?

Is it a billing meter or a monitoring meter? The two look similar on a parts list and behave differently in a dispute.

Does the meter serial on the device match the serial on the paperwork you hold?

If a current transformer is used, is the ratio documented on site and not only in the installer's file?

Who is authorised to break the seal, and what happens to the unit's billing status while it is open?

On older units this is where retrofit economics usually turn. A meter swap is a small part, but it brings resealing and re-documentation with it, and those costs do not scale down with the age of the cabinet.

Step four: check the software path — version, authority, horizon

Three questions per unit, all answerable from the back-end and one support email:

Version. Which protocol version is the controller actually running? Not what the datasheet lists as supported — what the unit reports when it connects.

Authority. Who signs the firmware, and can an update be applied remotely, or does it need a site visit with a laptop and a technician?

Horizon. What is the end-of-support date for this controller, in writing, and does it run beyond the period over which you plan to recover the retrofit cost?

Cyber requirements attach to a product's lifecycle rather than to a single test event. A unit that passed assessment years ago but receives no security updates for the rest of its service life is a growing liability, not a fixed one. The same logic applies to Plug & Charge readiness: part of the handshake lives in the vehicle and part in the charger's controller. If the controller cannot run the newer stack, no back-end change completes the sequence.

Step five: match the paperwork to the unit, not to the brochure

Four mismatches come up often enough to check every time. The declaration covers a variant with a different connector configuration from the one installed. The certificate holder is a different legal entity from the company that sold you the unit. The document is a photographed page with no annex listing covered variants. The test reports describe a component rather than the assembled product.

Ask for the full declaration with its annex, the variant list, the holder's legal name and the issue date relative to your serial numbers. If the supplier cannot produce that set, record the unit as undocumented and move on — an argument at this stage costs more than a line in the register. Documentation gaps are, in practice, the cheapest findings to close and the most expensive ones to ignore.

Step six: triage — retrofit, replace, or leave alone

With the register filled in, most findings fall into one of a few patterns. The middle column matters more than the last: a retrofit is only real if a certified variant exists that contains the missing function.

Where a site is being expanded at the same time, it is worth pricing the two paths side by side rather than retrofit by retrofit. A new high-power terminal can carry a new bay while older 120 kW and 180 kW units stay in service on lower-traffic positions — often cheaper than replacing everything in one budget year.

The three questions that decide the budget

Is the missing function available as a certified field-replaceable part? If the answer is a variant the manufacturer has assessed and declares, the retrofit is real. If it is a third-party module with no reissued declaration, you are buying hardware and inheriting the paperwork risk.

Is the platform supported beyond the payback period? A written end-of-support date is the only answer that counts. Without one, every unit of retrofit spend sits on a clock you cannot see.

Does the site have enough service life left to absorb it? Concessions, leases and site contracts end. A retrofit that outlives the site agreement is a cost with no revenue attached.

What operators tend to get wrong

Assuming a firmware update closes a hardware gap. It changes behaviour; it cannot add a card reader, a secure element or a billing meter.

Accepting a certificate screenshot as evidence. The variant list and the holder's legal name are the parts that get checked, and the parts a screenshot omits.

Treating retrofit and renewal as separate budgets. That is how a site ends up with new terminals at the front and unsupported cabinets at the back, on the same supply.

None of this needs a consultant. It needs one spreadsheet, a walk around the site with a torch, and the willingness to write “undocumented” in a cell instead of guessing. The documentation set shipped with a unit is the reference point for the paperwork column; if the file you hold does not match it, the gap is the finding. For sites inside a public network, the same register feeds the public fast-charging layout and the next round of fast-charging station planning. Two documents, one source of truth.